arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1909.01377cs.LGstat.ML
收藏

深度均衡模型

Deep Equilibrium Models

Carnegie Mellon University(卡内基梅隆大学);Bosch Center for AI(博世人工智能中心);Intel Labs(英特尔研究院)

Shaojie Bai, J. Zico Kolter, Vladlen Koltun

摘要:

我们提出了一种新的方法来建模序列数据:深度均衡模型(DEQ)。出于观察到许多现有的深层序列模型的隐藏层收敛到一些固定点,我们提出了DEQ方法,直接找到这些平衡点,通过寻根。这种方法相当于运行一个无限深度(权重绑定)的前馈网络,但具有显着的优势,我们可以使用隐式微分分析反向传播通过平衡点。使用这种方法,这些网络中的训练和预测只需要恒定的内存,而不管网络的有效“深度”如何。我们演示了DEQs如何应用于两个最先进的深度序列模型:自注意Transformers和网格网络。在大规模语言建模任务中,例如WikiText-103基准测试,我们表明DEQs 1)通常可以提高这些最先进的模型的性能(对于类似的参数计数); 2)与现有模型具有类似的计算要求; 3)极大地减少内存消耗(通常是训练大型序列模型的瓶颈),在我们的实验中显示了高达88%的内存减少。该代码可在https://github.com/locuslab/deq上获得。

英文摘要:

We present a new approach to modeling sequential data: the deep equilibrium model (DEQ). Motivated by an observation that the hidden layers of many existing deep sequence models converge towards some fixed point, we propose the DEQ approach that directly finds these equilibrium points via root-finding. Such a method is equivalent to running an infinite depth (weight-tied) feedforward network, but has the notable advantage that we can analytically backpropagate through the equilibrium point using implicit differentiation. Using this approach, training and prediction in these networks require only constant memory, regardless of the effective "depth" of the network. We demonstrate how DEQs can be applied to two state-of-the-art deep sequence models: self-attention transformers and trellis networks. On large-scale language modeling tasks, such as the WikiText-103 benchmark, we show that DEQs 1) often improve performance over these state-of-the-art models (for similar parameter counts); 2) have similar computational requirements to existing models; and 3) vastly reduce memory consumption (often the bottleneck for training large sequence models), demonstrating an up-to 88% memory reduction in our experiments. The code is available at https://github.com/locuslab/deq .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑